Mechanical probes of various sizes and shapes were used to determine thresholds for the perception of pressure, sharpness, and pain on the human finger. As force increased, perception changed from dull pressure to sharp pressure to sharp pain. With the smallest probe (0.01 mm2), sharpness threshold was very close to pressure threshold. As probe size increased, sharpness and pain threshold expressed in terms of force) increased in proportion to probe circumference (not probe area), whereas pressure threshold increased relatively little. Pain and sharpness thresholds also increased as probe angle became obtuse. There was a statistically significant increase in both thresholds with a probe angle change of 15 degrees. Thus, both size and shape are necessary to describe a mechanical stimulus adequately, and pressure (force/area) is not a sufficient metric for pain studies. Thresholds varied at different skin sites on the finger. The dorsal surface had lower thresholds than the volar surface, but the difference between the two areas was not always statistically significant. The compliance of the skin (e.g., the amount of indentation produced by a given force) exhibited no relation to sharpness or pain threshold, whether considered within subjects at various skin sites, or across subjects at the same skin site. Comparison of the perceptual thresholds with the thresholds for nociceptors determined in electrophysiological studies indicates that the sensation of nonpainful sharpness is likely to be mediated by nociceptors. Furthermore, considerably more than threshold activation of nociceptors is necessary for normal pain perception.